Getting it into your agent
This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.
/plugin marketplace add cogni-work/insight-wave/plugin install cogni-publishingWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/cogni-work/insight-wave/text-to-narrative)<a href="https://agentmods.dev/skills/cogni-work/insight-wave/text-to-narrative"><img src="https://agentmods.dev/badge/skills/cogni-work/insight-wave/text-to-narrative/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/cogni-work/insight-wave/text-to-narrative"><img src="https://agentmods.dev/badge/skills/cogni-work/insight-wave/text-to-narrative.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00182 | $0.06893 |
| Opus 5.5 | $0.00073 | $0.02757 |
| Sonnet 5.5 | $0.00036 | $0.01379 |
| Haiku 4.5 | $0.00018 | $0.00689 |
Grade A, and why
text-to-narrative scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 1mo ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 244 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Text to Narrative to Design Brief
Transform input markdown into a structured executive narrative using one of the bundled story arcs, then condense that narrative into one frozen design-brief.md for publishing or an optional Claude Design handoff. Narrative length is controlled by --target-length (default 1,675 words); section lengths are proportions of the total, so the arc's rhetorical balance survives at any scale. Each arc is one contract file that maps evidence to four elements, names the techniques that strengthen each, and states the arc's own validation rules. The brief's length is controlled by the target's density ceilings, and its copy is frozen: it selects from the narrative and never rewrites it.
Use this for: research syntheses, analyses or structured findings that need to become an executive narrative and a render-ready design brief in one run; a finished arc narrative that needs only the brief.
Not for: polishing arbitrary business documents (use copywriter); laying out or rendering the frozen brief (use design-compose and design-render); raw research (the cogni-knowledge pipeline).
Architectural model
The skill is a seven-phase pipeline: resolve context, select and validate one arc contract, draft four evidence-grounded elements, assemble and validate the narrative, freeze exact narrative spans into a design brief, then route that brief through publishing normalize → design-compose → design-render. ${CLAUDE_PLUGIN_ROOT}/references/arc-registry.md, the selected arc-{id}.md, execution-brief.md, validation.md, density-ceilings.md, and design-brief-template.md are the runtime authorities. Copy and order freeze at Phase 7; only presentation may change afterward. Claude Design is an optional provider route.
Parameters
| Parameter | Required | Description |
|---|---|---|
--source-path |
Yes | Directory containing input .md files, or path to a single .md file. A single file whose frontmatter carries both arc_id and word_count is a finished narrative: Phases 0-6 are skipped and only the brief is built from it |
--target |
No | Local publishing output: slides (default), document, infographic or web. Selects density ceilings, unit grammar and Rendering Contract wording; the same brief can optionally be handed to Claude Design |
--arc-id |
No | Explicit arc selection; overrides auto-detection |
--language |
No | Output language: en (default) or de. Fallback chain: explicit parameter > project metadata > workspace preference (.workspace-config.json) > content detection > en |
--output-path |
No | Narrative file path; defaults to insight-summary.md in the source directory |
--brief-path |
No | Design brief path; defaults to design-brief.md in the source directory |
--max-units |
No | Upper bound on brief units: slides, infographic blocks or web sections, lowering the target's own ceiling. Ignored on document, whose four sections are fixed — the checker's envelope says so. Default: the target's ceiling |
--theme-path |
No | Absolute path to a theme.md, recorded in the brief verbatim for the local renderer. Never prompted for; a theme is attached only when none is configured |
--project-path |
No | Research/knowledge project root; enables arc inheritance from the project's .metadata/ (Phase 1 step 8) and loading entity data beyond the source path. When omitted, Phase 1 step 8 probes <source-path>/.. and <source-path>/../.. |
--research-question |
No | Original research question, used for the subtitle and the opening |
--target-length |
No | Target total word count of the narrative (e.g., 2500). The acceptable range is ±15%. Default: 1675 (1,424-1,926 words). Recommended: 800-4,000 — outside that range the arc's proportions stop scaling well |
--content-map |
No | YAML map of content category keys to file/directory paths for additional context |
--audience |
No | Who the narrative is for. Default: senior business decision-makers. Feeds Pass 3 (vocabulary, acronym expansion, explanation depth) together with the inferred knowledge level |
--purpose |
No | The decision the narrative must serve. Default: understand the evidence and its strategic implications. Feeds Pass 2 (emphasis, close) and the Phase 5 TL;DR synthesis |
--perspective |
No | Whose voice the narrative speaks in. Default: neutral analyst. Feeds Pass 2 (pronouns, ownership) |
--geography |
No | Comma-separated BCP 47 region or established market codes already present in source/project metadata (for example de, fr, eu, global), never free text. Default: source-defined scope. Feeds Pass 1 evidence priority without requiring another plugin |
--interactive |
No | Whether the skill may pause for user input. true or false. Default: true. When false, skip all AskUserQuestion calls — there are two sites: the Phase 0 materiality clarification (takes the default instead) and the Phase 2 arc confirmation (keeps its top-ranked arc and its detection_reason and continues straight into Phase 3 with no prompt). Phase 7 has no prompt in either mode. Any value other than false is treated as true, so a malformed value fails safe toward the interactive default |
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 1mo ago First seen · 244 lines · 182 tokens per session scan A 5734393ab4b9
text-to-narrative is a skill published in the GitHub repository cogni-work/insight-wave (13 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 182 tokens to every session and 6,893 once invoked, about $0.0007 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-17.
Other skills, from other repositories
webgl-holographic-foil
A self-contained WebGL2 hero: thin-film interference over a crushed-foil surface whose palette shifts with the viewing angle; move the cursor to tilt the film.
html-ppt-hermes-cyber-terminal
OpenDesign + BYOK: choosing and wiring your own model, hands-on — cost, quality, and the routing decision. Built as a decision-grade AI literacy deck for engineers, IT, applied-AI teams.
html-ppt-taste-brutalist
16:9 HTML deck in tactical-telemetry / CRT-terminal taste. Deactivated-CRT charcoal slides, white-phosphor monospace, hazard-red accent, scanline overlay, ASCII syntax, density over decoration. Distilled from Leonxlnx/taste-skill brutalist-skill (Tactical Telemetry mode).
visual-ralph
Visual Ralph orchestration for frontend UI from generated references, static references, or live URL targets, using $ultragoal with built-in visual verdict and pixel-diff evidence until the implementation matches and leaves a reproducible design system.
glance-canonical-layouts
Use this skill to create home screen widgets using Jetpack Glance Canonical Layouts from the Android Platform Samples project, or when asked to implement standard widget layout patterns (including Check List, Image Grid, Snap Scrolling, and Standard Toolbar).
accessibility
Consolidated accessibility skill entrypoint for WCAG 2.2, ARIA Authoring Practices, cognitive accessibility, Section 508, EN 301 549, design intent verification, and the Accessibility Planner workflow.